Yuxun Tang

Here is the PR for audio resynthesis in discrete token: 1) We extend hubert_voc1 to token_voc1 and it can handle more models token; 2) We add f0 for training and...

[SVS2]: Discrete Singing Voice Synthesis

4

## What? PR for discrete singing voice synthesis (SVS2) 1. add new pipeline for SVS2 in Template 2. add recipe opencpop/svs2 3. update versa evaluation in svs2 4. support multi...

[SVS]: fix: fix visinger2 inference

1

## What did you change? fix visinger2 inference: support phoneme duration inference. --- ## Why did you make this change? fix bug. --- ## Is your PR small enough? yes...